Statistical Theory and Methods

Bayesian Robustness: One common critique of Bayesian statistics is its subjectivity, which arises through the necessity of specifying a prior distribution. In response to this critique, Bayesian statisticians sometimes analyze the sensitivity of their conclusions to the choice of prior. But conclusions can also be sensitive to the choice of sampling model. That was the subject of my Ph.D. dissertation.

Bayesian Nonparametrics: A few years after my Ph.D., Bill Sudderth introduced me to Polya trees. I spent several years investigating their properties and their utility for Bayesian nonparametric modelling.

Spatial Statistics: My next research interest was Markov random fields. I used them for density estimation and for investigating climate change in the ocean. I also studied the connections between Markov random fields and dynamic state-space models.

Statistical Applications
Ecology and Environment: I have had many fruitful and rewarding collaborations with ecologists and environmental scientists. Among them were a project to model the mortality of tree seedlings, the FACE experiment at Duke, and a long standing collaboration with Susan Lozier, an oceanographer, to investigate climate change in the ocean.

Neurobiology: I have been working with neurobiology to colleages to model spike trains, and the dependence among them, and optical images of the brain surface